WO2023048502A1 - 심전도를 기초로 갑상선 기능 장애를 진단하는 방법, 프로그램 및 장치 - Google Patents
심전도를 기초로 갑상선 기능 장애를 진단하는 방법, 프로그램 및 장치 Download PDFInfo
- Publication number
- WO2023048502A1 WO2023048502A1 PCT/KR2022/014258 KR2022014258W WO2023048502A1 WO 2023048502 A1 WO2023048502 A1 WO 2023048502A1 KR 2022014258 W KR2022014258 W KR 2022014258W WO 2023048502 A1 WO2023048502 A1 WO 2023048502A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- neural network
- network model
- electrocardiogram
- thyroid
- electrocardiogram data
- Prior art date
Links
- 230000006016 thyroid dysfunction Effects 0.000 title claims abstract description 59
- 238000000034 method Methods 0.000 title claims abstract description 53
- 238000003062 neural network model Methods 0.000 claims abstract description 106
- 238000013528 artificial neural network Methods 0.000 claims abstract description 30
- 210000001685 thyroid gland Anatomy 0.000 claims abstract description 18
- 206010020850 Hyperthyroidism Diseases 0.000 claims description 31
- 230000006870 function Effects 0.000 claims description 23
- 208000037170 Delayed Emergence from Anesthesia Diseases 0.000 claims description 11
- 208000001871 Tachycardia Diseases 0.000 claims description 7
- 238000004590 computer program Methods 0.000 claims description 7
- 230000006794 tachycardia Effects 0.000 claims description 7
- 102000011923 Thyrotropin Human genes 0.000 claims description 4
- 108010061174 Thyrotropin Proteins 0.000 claims description 4
- 238000004891 communication Methods 0.000 description 13
- 230000008569 process Effects 0.000 description 12
- 238000010586 diagram Methods 0.000 description 11
- 238000012545 processing Methods 0.000 description 9
- 238000010200 validation analysis Methods 0.000 description 9
- 238000012795 verification Methods 0.000 description 9
- 238000013103 analytical ultracentrifugation Methods 0.000 description 7
- 238000009601 thyroid function test Methods 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 6
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 4
- 201000010099 disease Diseases 0.000 description 4
- 238000011160 research Methods 0.000 description 4
- 238000012549 training Methods 0.000 description 4
- 210000000748 cardiovascular system Anatomy 0.000 description 3
- 230000008602 contraction Effects 0.000 description 3
- 238000013527 convolutional neural network Methods 0.000 description 3
- 238000003745 diagnosis Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 238000005259 measurement Methods 0.000 description 3
- 238000010206 sensitivity analysis Methods 0.000 description 3
- 208000024172 Cardiovascular disease Diseases 0.000 description 2
- 241000282412 Homo Species 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 210000000038 chest Anatomy 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000002565 electrocardiography Methods 0.000 description 2
- 230000004217 heart function Effects 0.000 description 2
- 238000001807 normal pulse voltammetry Methods 0.000 description 2
- 238000011056 performance test Methods 0.000 description 2
- 230000035945 sensitivity Effects 0.000 description 2
- 208000024891 symptom Diseases 0.000 description 2
- 239000005495 thyroid hormone Substances 0.000 description 2
- 229940036555 thyroid hormone Drugs 0.000 description 2
- 230000002861 ventricular Effects 0.000 description 2
- XUIIKFGFIJCVMT-GFCCVEGCSA-N D-thyroxine Chemical compound IC1=CC(C[C@@H](N)C(O)=O)=CC(I)=C1OC1=CC(I)=C(O)C(I)=C1 XUIIKFGFIJCVMT-GFCCVEGCSA-N 0.000 description 1
- 238000000585 Mann–Whitney U test Methods 0.000 description 1
- AUYYCJSJGJYCDS-LBPRGKRZSA-N Thyrolar Chemical class IC1=CC(C[C@H](N)C(O)=O)=CC(I)=C1OC1=CC=C(O)C(I)=C1 AUYYCJSJGJYCDS-LBPRGKRZSA-N 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000002567 autonomic effect Effects 0.000 description 1
- 230000007211 cardiovascular event Effects 0.000 description 1
- 238000000546 chi-square test Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000002526 effect on cardiovascular system Effects 0.000 description 1
- 210000001174 endocardium Anatomy 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 208000019622 heart disease Diseases 0.000 description 1
- 208000003532 hypothyroidism Diseases 0.000 description 1
- 230000002989 hypothyroidism Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 210000005240 left ventricle Anatomy 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 230000001404 mediated effect Effects 0.000 description 1
- 230000028161 membrane depolarization Effects 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000004660 morphological change Effects 0.000 description 1
- 210000004165 myocardium Anatomy 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000011176 pooling Methods 0.000 description 1
- 210000003742 purkinje fiber Anatomy 0.000 description 1
- 230000004213 regulation of atrial cardiomyocyte membrane depolarization Effects 0.000 description 1
- 230000034225 regulation of ventricular cardiomyocyte membrane depolarization Effects 0.000 description 1
- 230000013577 regulation of ventricular cardiomyocyte membrane repolarization Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 210000005245 right atrium Anatomy 0.000 description 1
- 210000005241 right ventricle Anatomy 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 210000001013 sinoatrial node Anatomy 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 210000000115 thoracic cavity Anatomy 0.000 description 1
- 229940034208 thyroxine Drugs 0.000 description 1
- XUIIKFGFIJCVMT-UHFFFAOYSA-N thyroxine-binding globulin Natural products IC1=CC(CC([NH3+])C([O-])=O)=CC(I)=C1OC1=CC(I)=C(O)C(I)=C1 XUIIKFGFIJCVMT-UHFFFAOYSA-N 0.000 description 1
- 238000012762 unpaired Student’s t-test Methods 0.000 description 1
- 230000002792 vascular Effects 0.000 description 1
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/318—Heart-related electrical modalities, e.g. electrocardiography [ECG]
- A61B5/346—Analysis of electrocardiograms
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
Definitions
- the present disclosure relates to a method for diagnosing thyroid dysfunction, and more specifically, to a method for diagnosing thyroid dysfunction using a neural network model based on an electrocardiogram.
- An electrocardiogram is a signal that can determine the presence or absence of a disease by measuring electrical signals generated from the heart to check for abnormalities in the conduction system from the heart to electrodes.
- the heartbeat which is the cause of the electrocardiogram, starts at the sinus node located in the right atrium. activate the ventricles
- the septum is fastest and the thin-walled right ventricle activates before the thick-walled left ventricle.
- the depolarization wave transmitted to the Purkinje fiber spreads from the endocardium to the epicardium like a wavefront in the myocardium, causing ventricular contraction.
- electrical impulses are conducted through the heart, causing the heart to contract about 60 to 100 times per minute. Each contraction is expressed as one heart rate.
- Such an electrocardiogram can be detected through a bipolar lead that records the potential difference between two parts and a unipolar lead that records the potential of the site where the electrode is attached.
- a bipolar lead that records the potential difference between two parts
- a unipolar lead that records the potential of the site where the electrode is attached.
- the electrical activity phase of the heart is largely divided into atrial depolarization, ventricular depolarization, and ventricular repolarization phases, and each of these phases is reflected in the form of several waves called P, Q, R, S, and T waves, as shown in FIG.
- electrocardiograms are measured with expensive measuring equipment and used as an auxiliary tool for measuring the patient's health condition.
- electrocardiogram measuring equipment displays only the measurement results, and the diagnosis is completely left to the doctor.
- the present disclosure has been made in response to the above-described background art, and a method for diagnosing thyroid dysfunction according to an embodiment of the present disclosure uses a neural network model to measure thyroid dysfunction for a target of electrocardiogram data based on electrocardiogram data. It is an object to provide a method for estimating the probability of occurrence.
- Diagnosis of thyroid dysfunction based on electrocardiogram performed by a computing device including at least one processor, according to an embodiment of the present disclosure for realizing the above object.
- a method comprising: obtaining electrocardiogram data; and estimating, based on the electrocardiogram data, an onset probability of thyroid dysfunction for a target of measuring the electrocardiogram data, using a pretrained neural network model, wherein the neural network model includes changes in thyroid function and electrocardiogram characteristics. It is possible to provide a method that is learned based on the correlation between the
- the neural network model may include a first sub-neural network model learned based on electrocardiogram data measured with 12 multi-leads.
- the neural network model may further include a second sub-neural network model learned based on at least one of six limb leads or six anterior chest leads.
- the neural network model may further include a third sub-neural network model learned based on electrocardiogram data measured with a single lead.
- the neural network model includes a neural network composed of a plurality of residual blocks, and the neural network composed of the residual blocks receives the electrocardiogram data and develops overt hyperthyroidism.
- a method for outputting probabilities can be provided.
- the overt hyperthyroidism may provide a method in which the level of urithyroxine is higher than a predetermined reference range or the thyroid stimulating hormone level is lower than the reference range.
- the neural network model includes a neural network corresponding to each of a plurality of leads of electrocardiogram data, and outputs of the neural networks are concatenated into one to derive the onset probability of thyroid dysfunction.
- the correlation between thyroid function and changes in electrocardiographic characteristics comprises at least one of: the frequency of tachycardia, the length of the QT interval, the direction of deviation of the P, R, and T waves, or the QRS duration.
- a method based on electrocardiogram characteristics may be provided.
- a method may be provided in which the probability of developing the thyroid dysfunction increases as the frequency of the tachycardia increases.
- a method may be provided in which the probability of developing thyroid dysfunction increases as the length of the QT interval increases.
- a method may be provided in which the occurrence probability of the thyroid dysfunction increases as the deviation directions of the P, R, and T waves go to the right.
- a method may be provided in which the occurrence probability of the thyroid dysfunction increases as the QRS duration becomes shorter.
- the step of estimating the onset probability of thyroid dysfunction for a subject to measure the electrocardiogram data based on the electrocardiogram data using a pretrained neural network model, the age and the electrocardiogram data together with the neural network model It is possible to provide a method that includes; inputting biological data including at least one of the sexes and estimating a probability of developing thyroid dysfunction for a target subject to measure the electrocardiogram data.
- a computer program stored on a computer readable storage medium when the computer program is executed on one or more processors, performs operations for diagnosing thyroid dysfunction based on an electrocardiogram. and the operations include obtaining electrocardiogram data; and estimating, based on the electrocardiogram data, an onset probability of thyroid dysfunction for a subject to measure the electrocardiogram data, using a pretrained neural network model, wherein the neural network model determines the relationship between thyroid function and changes in electrocardiogram characteristics. It is possible to provide a computer program, which is learned based on the correlation.
- a computing device for diagnosing thyroid dysfunction based on an electrocardiogram comprising: a processor including at least one core; and a memory including program codes executable by the processor, wherein the processor acquires electrocardiogram data according to execution of the program code, and uses a pre-trained neural network model, Based on the electrocardiogram data, a probability of occurrence of thyroid dysfunction is estimated for a target subject to measure the electrocardiogram data, and the neural network model is learned based on a correlation between changes in thyroid function and electrocardiogram characteristics.
- a method for diagnosing thyroid dysfunction may provide a method of estimating the onset probability of thyroid dysfunction for a target of electrocardiogram measurement based on electrocardiogram data using a neural network model.
- FIG. 1 is a diagram showing an electrocardiogram signal according to the present disclosure.
- FIG. 2 is a block diagram of a computing device according to one embodiment of the present disclosure.
- FIG. 3 is a flowchart illustrating a method of diagnosing thyroid dysfunction based on an electrocardiogram according to an embodiment of the present disclosure.
- FIG. 4 it is a diagram showing the structure of a neural network model according to an embodiment of the present disclosure.
- FIG. 5 is a diagram showing a verification research process of a neural network model according to an embodiment of the present disclosure.
- FIG. 6 is a diagram showing performance test results of a neural network model according to an embodiment of the present disclosure.
- FIG. 7 is a diagram showing electrocardiogram analysis results of subgroups classified by gender and age according to an embodiment of the present disclosure.
- x employs a or b should be understood to mean one of the natural inclusive substitutions.
- x employs a or b means that x employs a, x employs b, or x employs a and a. It can be interpreted as any one of the cases in which both of b are used.
- nth (n is a natural number) used in the present disclosure can be understood as an expression used to distinguish the components of the present disclosure from each other according to a predetermined criterion such as a functional point of view, a structural point of view, or explanatory convenience. there is.
- components performing different functional roles in the present disclosure may be classified as first components or second components.
- components that are substantially the same within the technical spirit of the present disclosure but should be distinguished for convenience of description may also be classified as first components or second components.
- acquisition used in the present disclosure is understood to mean not only receiving data through a wired/wireless communication network with an external device or system, but also generating data in an on-device form. It can be.
- module refers to a computer-related entity, firmware, software or part thereof, hardware or part thereof , It can be understood as a term referring to an independent functional unit that processes computing resources, such as a combination of software and hardware.
- a “module” or “unit” may be a unit composed of a single element or a unit expressed as a combination or set of a plurality of elements.
- a “module” or “unit” is a hardware element or set thereof of a computing device, an application program that performs a specific function of software, a process implemented through software execution, or a program. It may refer to a set of instructions for execution.
- a “module” or “unit” may refer to a computing device constituting a system or an application executed in the computing device.
- the concept of “module” or “unit” may be defined in various ways within a range understandable by those skilled in the art based on the content of the present disclosure.
- model used in this disclosure refers to a system implemented using mathematical concepts and language to solve a specific problem, a set of software units to solve a specific problem, or a process to solve a specific problem. It can be understood as an abstract model for a process.
- a neural network “model” may refer to an overall system implemented as a neural network having problem-solving capabilities through learning. At this time, the neural network may have problem solving ability by optimizing parameters connecting nodes or neurons through learning.
- a neural network "model” may include a single neural network or may include a neural network set in which a plurality of neural networks are combined.
- Data used in the present disclosure may include “image”, signals, and the like.
- image used in this disclosure may refer to multidimensional data composed of discrete image elements.
- image can be understood as a term referring to a digital representation of an object that is visible to the human eye.
- image may refer to multidimensional data composed of elements corresponding to pixels in a 2D image.
- Image may refer to multidimensional data composed of elements corresponding to voxels in a 3D image.
- block used in the present disclosure may be understood as a set of components classified based on various criteria such as type and function. Accordingly, a configuration classified as one “block” may be variously changed according to a criterion.
- a neural network “block” may be understood as a neural network set comprising at least one neural network. In this case, it may be assumed that the neural networks included in the neural network "block” perform the same specific operation. Explanations of the foregoing terms are intended to facilitate understanding of the present disclosure. Therefore, it should be noted that, when the above terms are not explicitly described as matters limiting the content of the present disclosure, the content of the present disclosure is not used in the sense of limiting the technical idea.
- FIG. 2 is a block diagram of a computing device according to an embodiment of the present disclosure.
- the computing device 100 may be a hardware device or part of a hardware device that performs comprehensive processing and calculation of data, or may be a software-based computing environment connected through a communication network.
- the computing device 100 may be a server that performs intensive data processing functions and shares resources, or may be a client that shares resources through interaction with the server.
- the computing device 100 may be a cloud system in which a plurality of servers and clients interact to comprehensively process data. Since the above description is only one example related to the type of the computing device 100, the type of the computing device 100 may be configured in various ways within a range understandable by those skilled in the art based on the contents of the present disclosure.
- a computing device 100 may include a processor 110, a memory 120, and a network unit 130. there is.
- the computing device 100 may include other configurations for implementing a computing environment. Also, only some of the components disclosed above may be included in the computing device 100 .
- the processor 110 may be understood as a structural unit including hardware and/or software for performing computing operations.
- the processor 110 may read a computer program and perform data processing for machine learning.
- the processor 110 may process input data processing for machine learning, feature extraction for machine learning, calculation of an error based on backpropagation, and the like.
- the processor 110 for performing such data processing includes a central processing unit (CPU), a general purpose graphics processing unit (GPGPU), a tensor processing unit (TPU), and on-demand It may include a semiconductor (application specific integrated circuit (ASICc)) or a field programmable gate array (FPGA). Since the above-described type of processor 110 is just one example, the type of processor 110 may be variously configured within a range understandable by those skilled in the art based on the content of the present disclosure.
- the processor 110 may train a neural network model for diagnosing thyroid dysfunction based on medical data.
- the processor 110 may train a neural network model to estimate the onset of hyperthyroidism, the degree of progression, and the like, based on electrocardiogram data and biological data including gender, age, and the like.
- the processor 110 may input electrocardiogram data and various kinds of biological data to the neural network model and train the neural network model so that the neural network model detects changes in the electrocardiogram according to the onset of hyperthyroidism.
- the neural network model may perform learning based on the correlation between thyroid function and electrocardiogram change. Correlation between thyroid function and electrocardiogram changes can be understood as information on the relationship between changes in thyroid function and morphological changes in electrocardiogram signals.
- the processor 110 may perform an operation representing at least one neural network block included in the neural network model during the learning process of the neural network model.
- the processor 110 may estimate whether thyroid dysfunction has occurred based on medical data using the neural network model generated through the above-described learning process.
- the processor 110 inputs electrocardiogram data and biological data including age and sex information into the neural network model learned through the above process to generate inference data representing the result of estimating the probability of developing thyroid dysfunction in humans.
- the processor 110 may input electrocardiogram data into a neural network model that has been trained, and may predict whether hyperthyroidism occurs or not, the degree of progression, and the like.
- the processor 110 can accurately predict the onset of thyroid dysfunction by effectively identifying subtle electrocardiogram changes that are difficult for humans to interpret through a neural network model for diagnosing thyroid dysfunction.
- the type of medical data and the output of the neural network model may be configured in various ways within a range understandable by those skilled in the art based on the contents of the present disclosure.
- the memory 120 may be understood as a unit including hardware and/or software for storing and managing data processed by the computing device 100 . That is, the memory 120 may store any type of data generated or determined by the processor 110 and any type of data received by the network unit 130 .
- the memory 120 may include a flash memory type, a hard disk type, a multimedia card micro type, a card type memory, and random access memory (RAM). ), SRAM (static random access memory), ROM (read-only memory), EEPROM (electrically erasable programmable read-only memory), PROM (programmable read-only memory), magnetic memory , a magnetic disk, and an optical disk may include at least one type of storage medium.
- the memory 120 may include a database system that controls and manages data in a predetermined system. Since the above-described type of memory 120 is just one example, the type of memory 120 may be configured in various ways within a range understandable by those skilled in the art based on the contents of the present disclosure.
- the memory 120 may organize and manage data necessary for the processor 110 to perform calculations, data combinations, program codes executable by the processor 110, and the like.
- the memory 120 may store medical data received through the network unit 130 to be described later.
- the memory 120 includes program codes for operating the neural network model to perform learning by receiving medical data, program codes for operating the neural network model to perform inference according to the purpose of use of the computing device 100 by receiving medical data, and Processing data generated as the program code is executed may be stored.
- the network unit 130 may be understood as a unit that transmits and receives data through any type of known wired/wireless communication system.
- the network unit 130 may include a local area network (LAN), wideband code division multiple access (WCDMA), long term evolution (LTE), and wireless broadband internet), 5th generation mobile communication (5g), ultra wide-band, zigbee, radio frequency (RF) communication, wireless LAN, wireless fidelity ), near field communication (NFC), or data transmission/reception may be performed using a wired/wireless communication system such as Bluetooth. Since the above-described communication systems are only examples, a wired/wireless communication system for data transmission and reception of the network unit 130 may be applied in various ways other than the above-described examples.
- the network unit 130 may receive data necessary for the processor 110 to perform an operation through wired/wireless communication with an arbitrary system or an arbitrary client.
- the network unit 130 may transmit data generated through the operation of the processor 110 through wired/wireless communication with an arbitrary system or an arbitrary client.
- the network unit 130 may receive medical data through communication with a database in a hospital environment, a cloud server that performs tasks such as standardization of medical data, or a computing device.
- the network unit 130 may transmit output data of the neural network model, intermediate data derived from the calculation process of the processor 110, and processed data through communication with the aforementioned database, server, or computing device.
- FIG. 3 is a flowchart illustrating a method of diagnosing thyroid dysfunction based on an electrocardiogram according to an embodiment of the present disclosure.
- a method of diagnosing thyroid dysfunction based on an electrocardiogram includes obtaining electrocardiogram data (S100). can be performed
- ECG data measured through an ECG measuring device may be directly acquired or acquired through network communication from the ECG measuring device.
- a step of estimating the onset probability of thyroid dysfunction for a target subject to measure the ECG data based on the ECG data using the pretrained neural network model (Ss110) may be performed.
- the estimating step (S110) may include the step of estimating the probability of developing thyroid dysfunction for the target of measurement of the electrocardiogram data by inputting the biological data including at least one of age and gender together with the electrocardiogram data into the neural network model.
- the neural network model may be learned based on a correlation between changes in thyroid function and electrocardiogram characteristics.
- the neural network model may be learned based on a correlation between changes in characteristics such as thyroid function and electrocardiogram, gender, and age.
- the neural network model may be learned based on the correlation between the onset and progress of hyperthyroidism and changes in electrocardiogram and other characteristics. Neural network models can be used to diagnose not only hyperthyroidism but also hypothyroidism and various thyroid dysfunctions.
- the neural network model may be learned based on an electrocardiogram measured with 12 leads obtained from electrodes of an electrocardiogram measuring device connected to the human body. For example, an electrocardiogram may be measured with 12 leads of 10 seconds in length and stored as 500 points per second.
- the neural network model may be learned based on partial information extracted from only 6 limb lead ECGs and a single lead I ECG among 12 lead ECGs.
- FIG. 4 it is a diagram showing the structure of a neural network model according to an embodiment of the present disclosure.
- a neural network model may include a neural network composed of a plurality of residual blocks.
- a neural network composed of residual blocks may be used to output an onset probability of overt hyperthyroidism by receiving electrocardiogram data.
- overt hyperthyroidism may be diagnosed as occurring when the level of vitreous thyroxine is higher than a predetermined reference range or the thyroid stimulating hormone level is lower than the reference range.
- the neural network model may have a structure of a resnet neural network using 6 residual blocks.
- Each residual block may include a convolutional neural network (CNN), batch normalization, a rectified linear unit (hereinafter, ReLU) function, and a dropout layer.
- the convolutional neural network is 1-dimensional and the size of the filter can be set to 21.
- the input length may be halved whenever three residual blocks are passed among a total of six residual blocks.
- Different neural networks may be applied to each lead of the ECG.
- average pooling may be applied in units of channels.
- the output of the neural networks can be concatenated to derive the probability of developing thyroid dysfunction.
- the neural network model may include a neural network corresponding to each of a plurality of leads of electrocardiogram data. That is, the neural network model may include individual neural networks into which electrocardiograms measured by individual leads are respectively input.
- the neural network model may include a first sub-neural network model learned based on electrocardiogram data measured with 12 multi-leads.
- the neural network model may further include a second sub-neural network model learned based on at least one of six limb leads or six anterior chest leads.
- the neural network model may further include a third sub-neural network model learned based on electrocardiogram data measured with a single lead.
- the neural network model may selectively use at least one of a first sub neural network model, a second sub neural network model, and a third sub neural network model according to the number of leads. Therefore, the neural network model can effectively predict the onset of thyroid dysfunction regardless of the number of leads.
- the neural network model uses all of the first sub-neural network model, the second sub-neural network model, and the third sub-neural network model and combines the outputs of each sub-model to determine the onset probability of thyroid dysfunction. can also be output. Through this combination, neural network models can increase the accuracy of predicting the onset of thyroid dysfunction.
- the performance of the neural network model was verified by comparing the probability calculated by the model with the presence or absence of hyperthyroidism in the internal/external verification data set. Verification was performed with reference to the area under the receiver operating characteristic curve (hereinafter, AUC). The cutoff point was identified using the youden j statistic in the training data set. Cutoff points were applied to calculate sensitivity, specificity, positive predictive value, and negative predictive value in the internal/external validation data set. The 95% confidence interval of AUC was calculated using sun&su's optimization of the de-long method.
- a sensitivity analysis was conducted by making subgroups according to age and gender. Gender was classified as male and female, and age was classified into less than 40 years old, 40 to less than 50 years of age, 50 to less than 60 years of age, 60 to less than 70 years of age, and 70 years of age or older.
- the external validation data set was extracted from patients who received a normal diagnosis in the first thyroid function test (TFT) and underwent subsequent thyroid function tests.
- the time interval between the first thyroid function test and the subsequent thyroid function test is 4 weeks or more.
- the cutoff point was determined using the Juden J statistic on the training data set. We used the Kaplan-Meier method to analyze the results over 36 months.
- FIG. 5 is a diagram showing a verification research process of a neural network model according to an embodiment of the present disclosure.
- the subjects of the verification study of the neural network model are 113,215 hospital A patients and 33,485 hospital B patients. 21 hospital A patients and 7 hospital B patients with missing clinical information or ECG data were excluded. A total of 2164 patients with hyperthyroidism were included.
- 139,521 ECG data measured from 90,554 patients in Hospital A were used.
- Internal validation used 34,810 ECG data from 518 patients in Hospital A.
- External verification 48,684 ECG data from 33,478 patients in Hospital B.
- the alternative hypothesis for p marked with ⁇ in Table 1 is that there is a difference between hyperthyroidism and overt hyperthyroidism.
- ⁇ The alternative hypothesis for the indicated p values is that there is a difference between hospital A (model development and internal validation data group) and hospital B (external validation group) for each variable. Gender, age, and incidence of hyperthyroidism showed statistically significant differences between hospitals. Patients with hyperthyroidism had more tachycardia and longer QT interval. Patients with hyperthyroidism showed deflection of the P-wave, R-wave, and T-wave axes to the right, and the QRS duration was short.
- the above-described correlation between changes in thyroid function and electrocardiogram characteristics includes at least one of the frequency of tachycardia, the length of the QT interval, the direction of deviation of the P, R, and T waves, or the QRS duration. It can be based on the characteristics of the electrocardiogram.
- the probability of developing thyroid dysfunction estimated by the neural network model may increase as the frequency of tachycardia increases.
- the probability of developing thyroid dysfunction may increase as the length of the QT interval increases.
- the probability of occurrence of thyroid dysfunction estimated by the neural network model may increase as the deviation directions of the P, R, and T waves move toward the right.
- the probability of developing thyroid dysfunction estimated by the neural network model may increase as the QRS duration is shortened.
- FIG. 6 is a diagram showing performance test results of a neural network model according to an embodiment of the present disclosure.
- AUC is the area under the receiver operating characteristic curve of the neural network model
- DLM is the neural network model
- ECG is the electrocardiogram
- NPV is the negative predictive value
- PPV is the positive predictive value
- SEN stands for sensitivity
- SPE stands for specificity.
- the AUCs of the neural network model using 12-lead ECG were 0.918 (0.909-0.927) and 0.897 (0.879-0.916), respectively.
- Sensitivity analysis confirmed the robustness of the neural network model according to gender and age. Those identified by the neural network model as high-risk patients showed a significant change in the incidence of hyperthyroidism (p ⁇ 0.01) compared to those identified as low-risk patients.
- the performance of the neural network model using 6-lead ECG and single-lead ECG can also be seen. Referring to [Table 2], the performance of all models in the sensitivity analysis for gender and age showed AUC values of 0.830 or higher.
- FIG. 7 is a diagram showing electrocardiogram analysis results of subgroups classified by gender and age according to an embodiment of the present disclosure.
- subgroup analysis is performed on 6,762 patients who were found to be normal in a thyroid function test after follow-up. The analysis was performed using the thyroid function test results. Of these, 24 patients developed hyperthyroidism. Subjects for subgroup analysis were divided into 4,749 high-risk and 2,013 low-risk groups according to the probability of developing hyperthyroidism output by the neural network model. It was confirmed that the high-risk group had a significantly higher risk of hyperthyroidism than the low-risk group (0.48% vs. 0.05%, p ⁇ 0.01).
- thyroid function is closely related to the cardiovascular system and can affect cardiac function, vascular resistance, cardiovascular autonomic control function, as well as the cardiovascular system.
- ventricular contraction (inotropy) and heart rate (chronotropy) with enhanced relaxation function due to thyroid hormone-mediated changes can affect the improvement of cardiac function. Signs and symptoms of thyroid dysfunction can be judged as a result of thyroid hormones affecting the heart and cardiovascular system.
- Thyroid dysfunction can be associated with increased cardiovascular disease morbidity and mortality.
- untreated hyperthyroidism was associated with a higher risk of cardiovascular disease than treated patients. Cardiovascular events increased with the duration of decreased thyroid-stimulating hormone levels in both treated and untreated cases of hyperthyroidism.
- the method for diagnosing thyroid dysfunction uses a neural network model, based on the electrocardiogram data, to measure the thyroid function of the target of electrocardiogram data. It was possible to estimate the probability of occurrence of the disorder.
- the method for diagnosing thyroid dysfunction has an effect of diagnosing hyperthyroidism using a neural network model based on information such as electrocardiogram, gender, and age.
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Medical Informatics (AREA)
- Public Health (AREA)
- General Health & Medical Sciences (AREA)
- Pathology (AREA)
- Heart & Thoracic Surgery (AREA)
- Cardiology (AREA)
- Physics & Mathematics (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- Veterinary Medicine (AREA)
- Epidemiology (AREA)
- Primary Health Care (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
Description
내부 검증 데이터 세트 (병원 a) n=113,194 |
외부 검증 데이터 세트 (병원 b) n=33,478 |
p‡ | |||||
특성 | 비(非) 갑상선 기능 항진증 | 갑상선 기능 항진증 | p† | 비(非) 갑상선 기능 항진증 | 갑상선 기능 항진증 | p† | p‡ |
연구대상, 명 (%) | 111,195 (98.2) | 1999 (1.8) | 33,313 (99.5) |
165 (0.5) | <0.001 | ||
나이, 세, 평균(sd) | 43.95 (13.41) | 40.64 (14.38) | <0.001 | 55.37 (15.48) | 51.51 (14.93) | 0.001 | <0.001 |
남성, 명, (%) | 56808 (51.1) | 638 (31.9) | <0.001 | 15960 (47.9) | 56 (33.9) | <0.001 | <0.001 |
심박수, bpm (%) | 66.47 (12.75) | 89.37 (19.61) | <0.001 | 70.89 (16.20) | 97.59 (24.19) | <0.001 | <0.001 |
PR간격, ms, 평균 (sd) | 158.46 (23.63) | 145.83 (25.51) | <0.001 | 166.97 (26.47) | 151.08 (27.45) | <0.001 | <0.001 |
QRS길이, ms, 평균 (sd) | 93.07 (12.00) | 86.24 (11.37) | <0.001 | 94.36 (15.02) | 89.67 (13.46) | <0.001 | <0.001 |
QTC, ms, 평균 (sd) | 418.46 (23.42) | 427.41 (32.29) | <0.001 | 433.44 (31.63) | 442.88 (31.23) | <0.001 | <0.001 |
P파의 축, 평균(sd) | 47.38 (23.68) | 50.37 (24.46) | <0.001 | 44.34 (27.65) | 48.93 (31.43) | 0.045 | <0.001 |
R파의 축, 평균(sd) | 49.42 (32.13) | 54.86 (28.28) | <0.001 | 40.51 (39.55) | 46.27 (33.08) | 0.062 | 0.002 |
T파의 축, 평균(sd) | 39.83 (23.53) | 46.32 (25.33) | <0.001 | 39.38 (38.21) | 47.95 (47.98) | 0.004 | <0.001 |
남성 | 여성 | |||||||||
나이 | AUC | SEN | SPE | PPV | NPV | AUC | SEN | SPE | PPV | NPV |
-39 | 0.919 (0.899-0.940) | 0.839 (0.793-0.884) | 0.883 (0.877-0.890) | 0.169 (0.148-0.190) | 0.995 (0.993-0.996) | 0.932 (0.913-0.951) | 0.841 (0.794-0.889) | 0.897 (0.891-0.903) | 0.173 (0.151-0.196) | 0.995 (0.994-0.997) |
40-49 | 0.893 (0.868-0.917) | 0.790 (0.736-0.844) | 0.872 (0.865-0.879) | 0.131 (0.113-0.149) | 0.994 (0.992-0.996) | 0.911 (0.887-0.934) | 0.809 (0.759-0.859) | 0.915 (0.910-0.920) | 0.166 (0.144-0.187) | 0.996 (0.994-0.997) |
50-59 | 0.880 (0.853-0.907) | 0.823 (0.768-0.877) | 0.803 (0.795-0.810) | 0.072 (0.061-0.083) | 0.996 (0.994-0.997) | 0.904 (0.877-0.931) | 0.812 (0.755-0.870) | 0.855 (0.848-0.862) | 0.099 (0.083-0.114) | 0.996 (0.994-0.997) |
60-69 | 0.830 (0.787-0.874) | 0.682 (0.595-0.769) | 0.876 (0.868-0.883) | 0.079 (0.061-0.096) | 0.994 (0.993-0.996) | 0.868 (0.829-0.908) | 0.754 (0.675-0.833) | 0.840 (0.831-0.850) | 0.084 (0.067-0.101) | 0.994 (0.992-0.996) |
70- | 0.873 (0.827-0.919) | 0.836 (0.751-0.921) | 0.825 (0.814-0.835) | 0.064 (0.049-0.080) | 0.997 (0.996-0.999) | 0.852 (0.804-0.900) | 0.957 (0.900-1.015) | 0.674 (0.663-0.685) | 0.019 (0.014-0.025) | 1.000 (0.999-1.000) |
Claims (15)
- 적어도 하나의 프로세서를 포함하는 컴퓨팅 장치에 의해 수행되는, 심전도를 기초로 갑상선 기능 장애(dysfunction)를 진단하는 방법으로서,심전도 데이터를 획득하는 단계; 및사전 학습된 신경망 모델을 사용하여, 상기 심전도 데이터를 기초로 상기 심전도 데이터의 측정 대상에 대한 갑상선 기능 장애의 발병 확률을 추정하는 단계를 포함하고,상기 신경망 모델은,갑상선 기능과 심전도 특성의 변화 간의 상관관계를 기초로 학습된 것인,방법.
- 제1항에 있어서,상기 신경망 모델은,12개의 다중 리드(lead)로 측정되는 심전도 데이터를 기초로 학습된 제1 서브 신경망 모델을 포함하는,방법.
- 제1항에 있어서,상기 신경망 모델은,6개의 림브(Limb) 리드, 혹은 6개의 전흉부 리드 중 적어도 하나를 기초로 학습된 제2 서브 신경망 모델을 포함하는,방법.
- 제1항에 있어서,상기 신경망 모델은,단일 리드로 측정되는 심전도 데이터를 기초로 학습된 제3 서브 신경망 모델을 포함하는,방법.
- 제1항에 있어서,상기 신경망 모델은,복수의 레지듀얼 블록들(Residual blocks)로 구성되는 신경망을 포함하고,상기 레지듀얼 블록들로 구성되는 신경망은,상기 심전도 데이터를 입력 받아 현성 갑상선 기능 항진증의 발병 확률을 출력하는,방법.
- 제5항에 있어서,상기 현성 갑상선 기능 항진증은,유리티록신 수치가 사전 결정된 기준 범위보다 높거나, 갑상선 자극 호르몬 수치가 기준 범위보다 낮은 경우인,방법.
- 제1항에 있어서,상기 신경망 모델은,심전도 데이터의 복수의 리드들 각각에 대응되는 신경망을 포함하고,상기 신경망들의 출력은,갑상선 기능 장애의 발병 확률을 도출하기 위해 하나로 연결(concatenation)되는,방법.
- 제1항에 있어서,갑상선 기능과 심전도 특성의 변화 간의 상관관계는,빈맥의 빈도, QT 간격(interval)의 길이, P파, R파 및 T파의 편위 방향, 또는 QRS 지속시간 중 적어도 하나를 포함하는 심전도 특성에 기반하는,방법.
- 제8항에 있어서,상기 갑상선 기능 장애의 발병 확률은,상기 빈맥의 빈도가 많을수록 높아지는,방법.
- 제8항에 있어서,상기 갑상선 기능 장애의 발병 확률은,상기 QT 간격의 길이가 길수록 높아지는,방법.
- 제8항에 있어서,상기 갑상선 기능 장애의 발병 확률은,상기 P파, R파 및 T파의 편위 방향이 우측을 향할수록 높아지는,방법.
- 제8항에 있어서,상기 갑상선 기능 장애의 발병 확률은,상기 QRS 지속시간이 짧을수록 높아지는,방법.
- 제1항에 있어서,사전 학습된 신경망 모델을 사용하여, 상기 심전도 데이터를 기초로 상기 심전도 데이터의 측정 대상에 대한 갑상선 기능 장애의 발병 확률을 추정하는 단계는,상기 신경망 모델로 상기 심전도 데이터와 함께 나이 및 성별 중 적어도 하나를 포함하는 생물학적 데이터를 입력하여, 상기 심전도 데이터의 측정 대상에 대한 갑상선 기능 장애의 발병 확률을 추정하는 단계; 를 포함하는,방법.
- 컴퓨터 판독가능 저장 매체 저장된 컴퓨터 프로그램(program)으로서, 상기 컴퓨터 프로그램은 하나 이상의 프로세서(processor)에서 실행되는 경우, 심전도를 기초로 하는 갑상선 기능 장애 진단을 위한 동작들을 수행하도록 하며,상기 동작들은, 심전도 데이터를 획득하는 동작; 및사전 학습된 신경망 모델을 사용하여, 상기 심전도 데이터를 기초로 상기 심전도 데이터의 측정 대상에 대한 갑상선 기능 장애의 발병 확률을 추정하는 동작을 포함하고,상기 신경망 모델은 갑상선 기능과 심전도 특성의 변화 간의 상관관계를 기초로 학습된 것인,컴퓨터 프로그램.
- 심전도를 기초로 하는 갑상선 기능 장애 진단을 위한 컴퓨팅 장치로서, 적어도 하나의 코어(core)를 포함하는 프로세서(processor); 및 상기 프로세서에서 실행 가능한 프로그램 코드(code)들을 포함하는 메모리(memory);를 포함하고,상기 프로세서는, 상기 프로그램 코드의 실행에 따라,심전도 데이터를 획득하고, 사전 학습된 신경망 모델을 사용하여, 상기 심전도 데이터를 기초로 상기 심전도 데이터의 측정 대상에 대한 갑상선 기능 장애의 발병 확률을 추정하며,상기 신경망 모델은 갑상선 기능과 심전도 특성의 변화 간의 상관관계를 기초로 학습된 것인, 장치.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202280062075.5A CN117940060A (zh) | 2021-09-25 | 2022-09-23 | 以心电图为基础诊断甲状腺功能障碍的方法、程序及装置 |
Applications Claiming Priority (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
KR10-2021-0126785 | 2021-09-25 | ||
KR20210126785 | 2021-09-25 | ||
KR1020220118440A KR20230044125A (ko) | 2021-09-25 | 2022-09-20 | 심전도를 기초로 갑상선 기능 장애를 진단하는 방법, 프로그램 및 장치 |
KR10-2022-0118440 | 2022-09-20 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2023048502A1 true WO2023048502A1 (ko) | 2023-03-30 |
Family
ID=85720941
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/KR2022/014258 WO2023048502A1 (ko) | 2021-09-25 | 2022-09-23 | 심전도를 기초로 갑상선 기능 장애를 진단하는 방법, 프로그램 및 장치 |
Country Status (1)
Country | Link |
---|---|
WO (1) | WO2023048502A1 (ko) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20060117546A (ko) * | 2005-05-11 | 2006-11-17 | 인하대학교 산학협력단 | 신경망을 이용한 심전도 기반의 심장질환 진단장치 및방법 |
KR20110011911A (ko) * | 2009-07-29 | 2011-02-09 | 한국과학기술원 | 생체신호 추정을 통한 질병검진 장치 및 방법 |
JP2020074940A (ja) * | 2018-11-08 | 2020-05-21 | 兆奇 胡 | 脈拍測定による健康状態評価システムとその作動方法 |
KR20210058274A (ko) * | 2019-11-14 | 2021-05-24 | 권준명 | 머신러닝을 기반으로 생성된 심전도표준데이터를 이용하여 사용자의 신체상태를 판단하는 심전도 측정 시스템 및 그 방법 |
-
2022
- 2022-09-23 WO PCT/KR2022/014258 patent/WO2023048502A1/ko active Application Filing
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20060117546A (ko) * | 2005-05-11 | 2006-11-17 | 인하대학교 산학협력단 | 신경망을 이용한 심전도 기반의 심장질환 진단장치 및방법 |
KR20110011911A (ko) * | 2009-07-29 | 2011-02-09 | 한국과학기술원 | 생체신호 추정을 통한 질병검진 장치 및 방법 |
JP2020074940A (ja) * | 2018-11-08 | 2020-05-21 | 兆奇 胡 | 脈拍測定による健康状態評価システムとその作動方法 |
KR20210058274A (ko) * | 2019-11-14 | 2021-05-24 | 권준명 | 머신러닝을 기반으로 생성된 심전도표준데이터를 이용하여 사용자의 신체상태를 판단하는 심전도 측정 시스템 및 그 방법 |
Non-Patent Citations (1)
Title |
---|
DEEPIKA M., KALAISELVI K.: "A Contemplative Analysis of Thyroid Disorders based on ECG Signal and Thyroid Test Measure", EAI ENDORSED TRANSACTIONS ON ENERGY WEB, vol. 8, no. 32, XP093053517, DOI: 10.4108/eai.1-7-2020.166006 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Murugesan et al. | Ecgnet: Deep network for arrhythmia classification | |
US20200335217A1 (en) | Methods and systems using mathematical analysis and machine learning to diagnose disease | |
US11617528B2 (en) | Systems and methods for reduced lead electrocardiogram diagnosis using deep neural networks and rule-based systems | |
CN109846474B (zh) | 心电图的处理方法及装置、心电图的远程处理方法及系统 | |
WO2022119155A1 (ko) | 설명 가능한 다중 심전도 부정맥 진단 장치 및 방법 | |
WO2023103156A1 (zh) | 一种基于心跳节律信号的房颤发生风险预测系统及其应用 | |
Hong et al. | Cardiolearn: a cloud deep learning service for cardiac disease detection from electrocardiogram | |
TWI688371B (zh) | 心房顫動信號型態擷取及輔助診斷智能裝置 | |
CN114343585A (zh) | 认知和行为障碍的预警方法、装置、设备及存储介质 | |
WO2023048502A1 (ko) | 심전도를 기초로 갑상선 기능 장애를 진단하는 방법, 프로그램 및 장치 | |
Rahman et al. | A real-time tunable ECG noise-aware system for IoT-enabled devices | |
WO2023063619A1 (ko) | 심전도를 기초로 좌심실 수축기 장애를 진단하는 방법, 프로그램 및 장치 | |
KR20230044125A (ko) | 심전도를 기초로 갑상선 기능 장애를 진단하는 방법, 프로그램 및 장치 | |
WO2024019584A1 (ko) | 심전도를 이용한 심근경색 진단 방법, 프로그램 및 장치 | |
CN117940060A (zh) | 以心电图为基础诊断甲状腺功能障碍的方法、程序及装置 | |
Saadi et al. | Low-cost ecg monitoring system with classification using deep learning | |
Laudato et al. | Combining rhythmic and morphological ECG features for automatic detection of atrial fibrillation: local and global prediction models | |
Rakhmatullov et al. | Technologies for optimization and digitalization of cardiological care | |
WO2023048486A1 (ko) | 심전도 신호의 오류 복원 방법, 프로그램 및 장치 | |
WO2024101759A1 (ko) | 심전도 데이터의 분석 방법, 프로그램 및 장치 | |
WO2024019575A1 (ko) | 심전도를 이용한 건강 상태의 예측 방법, 프로그램 및 장치 | |
WO2024014838A1 (ko) | 심전도 판독에 기반한 시각화 콘텐츠를 제공하는 방법, 프로그램 및 장치 | |
WO2024014844A1 (ko) | 심전도에 기반하여 체성분을 측정하는 방법 및 그 장치 | |
WO2024019523A1 (ko) | 심전도를 이용한 건강 상태의 예측 방법, 프로그램 및 장치 | |
WO2024049193A1 (ko) | 심전도 데이터의 거래 서비스를 제공하는 방법, 프로그램 및 장치 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 22873206 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 202280062075.5 Country of ref document: CN |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2022873206 Country of ref document: EP |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
ENP | Entry into the national phase |
Ref document number: 2022873206 Country of ref document: EP Effective date: 20240425 |